As I've pointed out on air and in my blog repeatedly over the years, so-called "Tornado Sirens" are NOT meant to warn you UNLESS you are outside... and that assumes they work properly and some or all of them don't fail.

They are not designed for you to hear them inside your home or a business, and they are not designed to wake you up if you're asleep! Many in the "forecasting community" even argue they should all be eliminated because they are going to get somebody killed, if they haven't already…because their existence makes many people wait to hear them before ducking and covering or rely on them instead of radio/TV. This is why so often someone in the path says "it struck without warning" but it is then shown with proof there was advanced warning, the person just did not hear the warning.

Doppler radar does a great job of detecting tornadoes, however it is not fool proof in determining if rotation in the storm clouds is making it to the ground as an actual tornado. For that reason, the false alarm rate on tornado warnings is still too high, very high. This is the reason that despite dozens of tornado warnings issued the other day, there was actually only one real tornado that we know of so far.

In fact, notice in the graphics with this blog post that the strong rotation indicating a possible tornado started way over in Alabama, it took quite some time for it to touch down as the killer tornado near the Bartow-Gordon County line. One thing I told our listners about the day before the storms and early that morning, was that the situation could be particularly sneaky and dangerous because the storms forward speed would be unusually fast-- decreasing warning time, and that there would not be a lot of lightning and thunder with the heavy rain, meaning mother nature would not give you the usual notice that a storm was approaching you. Valuable information you didn't get anywhere else.

The detection rate for tornadoes on Doppler is very high, sadly so is the false alarm rate. You may die if you rely on and wait for so-called "Tornado Sirens".  I did a screen grab of the Atlanta hourly weather forecast off a well know weather company website for that day, it showed cloudy with a 10% chance of rain! Automated forecasts on the web or mobile devices are robots with no meteorologist involved. This can lead you astray in a big bad way if you rely on them in changeable weather. RADIO is the original social media for weather! Radio is the original web source.

A TEAM OF METEOROLOGISTS FROM THE NATIONAL WEATHER SERVICE IN
PEACHTREE CITY GEORGIA SURVEYED THE DAMAGE CAUSED BY A SUPERCELL
THUNDERSTORM THAT TRACKED ACROSS NORTHWEST BARTOW AND CENTRAL GORDON
COUNTIES FROM APPROXIMATELY 1115 AM TO 1155 AM ON WEDNESDAY...
JANUARY 30 2013. RESULTS FROM THE SURVEY INCLUDE THE FOLLOWING.

.ADAIRSVILLE TORNADO...

RATING:                 HIGH END EF-3
ESTIMATED PEAK WIND:    160 MPH
PATH LENGTH /STATUTE/:  21.8 MILES
PATH WIDTH /MAXIMUM/:   900 YARDS
FATALITIES:             1
INJURIES:               17 /9 IN BARTOW...8 IN GORDON/

START DATE:             JAN 30 2013
START TIME:             1112 AM EST
START LOCATION:         2.4 MI SW ADAIRSVILLE/BARTOW COUNTY START
LAT/LON:                34.3411N/-84.9527W

END DATE:               JAN 30 2013
END TIME:               1143 AM EST
END LOCATION: 14.3 MI NE CALHOUN OR NEAR OAKMAN/GORDON COUNTY
END LAT/LON: 34.5981N/-84.7313W

IN BARTOW COUNTY...AT LEAST 95 STRUCTURES WERE DAMAGED...WITH 31
DESTROYED... 17 SUSTAINING MAJOR DAMAGE...AND 47 WITH MINOR DAMAGE.
MOST OF THE DAMAGE WAS A RESULT OF TREES FALLING ON HOMES IN
DOWNTOWN ADAIRSVILLE. IN GORDON COUNTY...268 HOMES STRUCTURES WERE
IMPACTED. OF THESE...30 WERE COMPLETELY DESTROYED...110 HAD MAJOR
DAMAGE... AND ANOTHER 70 HAD MINOR DAMAGE. 202 OF THESE WERE SINGLE
FAMILY HOMES AND 66 WERE MOBILE HOMES.

EF SCALE: THE ENHANCED FUJITA SCALE CLASSIFIES TORNADOES INTO
THE FOLLOWING CATEGORIES.

EF0...WEAK......65 TO 85 MPH
EF1...WEAK......86 TO 110 MPH
EF2...STRONG....111 TO 135 MPH
EF3...STRONG....136 TO 165 MPH
EF4...VIOLENT...166 TO 200MPH
EF5...VIOLENT...>200MPH

Event Summary from Atlanta National Weather Service:

In the days leading up to this event, a strong ridge aloft developed, which along with surface winds from the south and southeast, led to near record warmth with temperatures on the 29th and 30th in the lower 70s to the north to near 80 for the southern tier of our forecast area.  Meanwhile, a deep and sharp upper level trough transitioned from the Great Plains states to the mid Mississippi Valley late on the 29th and into the morning of the 30th.  With it came much colder air for the Upper Plains with temperatures well below freezing and the freezing line itself making it as far south as Oklahoma. This clash of temperatures set the stage for a multi-state long-duration severe weather event that would produce widespread wind damage and several significant tornadoes across the local area. Here are some images of the setup for this event.

Surface analysis shown on the morning of the 30th showing Low pressure developing along the front over Southern Mississippi. Note the warm air ahead of the front with temperatures in the 60s in Georgia at 7am	This is the satellite imagery for the same timeframe indicating the wide scope of influence this storm system would have with storms from the Gulf Coast through Canada and snow over the Northern Plains.

Forecasting the Event

By most measures, this was a well forecast event by the computer models and well advertised with mention of the potential for severe storms up to a week in advance while the North Georgia ice storm was ongoing.

Graphicast showing initial advertisement of severe potential 4 days out.	Graphicast the day of the event detailing the timing of the line as it progressed through the area.

Like most significant events, this one was not without its set of unique challenges and this one included determining how soon line of thunderstorms emanating from the Mississippi Valley would decrease in intensity. Computer models were adamant about weakening the line as it progressed east but the question was would this result in less severe weather locally? Let's look at some images from the event.

SREF computer model showing high severe thunderstorm probability over Mississippi.	SREF computer model showing much less probabilities as the line shifts east into Georgia.

So how did the forecasts turn out? Well in actuality the models were right on target in depicting the decreasing intensity to the storms as they moved through the area. Look at these images showing the lightning in pink as the line moves through Mississippi.

Contrast this to the same map as the storms moved through Georgia and you can see the lightning did in fact decrease significantly around mid morning.

So where did all the severe weather come from if there was a weakening line of storms? Well for this event, there was one parameter that stood out as being extremely unusual in terms of strength. We call this an anomaly and the anomaly for low level winds created reason for concern early on. The 925mb winds were forecast to be between 55 and 65 mph for this event which means we were expecting very strong winds within a few thousand feet off the ground.  Being so close to the ground, these winds would be able to reach the surface and were largely independent of the individual storm strength, meaning severe winds would be possible regardless of the storm intensity.  In addition, despite the weakening trend, we still had healthy storms as the line moved into Western Georgia with sufficient lift to take advantage low level wind shear an produce the unfortunate Adairsville tornado.